American Journal of Applied Mathematics

2015; 3(1-1): 30-42

Published online December 25, 2014 (http://www.sciencepublishinggroup.com/j/ajam)
doi: 10.11648/j.ajam.s.2015030101.14
ISSN: 2330-0043 (Print); ISSN: 2330-006X (Online)

Numerical solutions of 2-D incompressible driven cavity

flow with wavy bottom surface
K. M. Salah Uddin1, Litan Kumar Saha2
1
Department of Management Information Systems, University of Dhaka, Dhaka, Bangladesh
2
Deparment of Applied Mathematics, University of Dhaka, Dhaka, Bangladesh

Email address:
salahuddin@du.ac.bd (K. M. S. Uddin), lksaha.math@gmail.com (L. K. Saha)

To cite this article:

K. M. Salah Uddin, Litan Kumar Saha. Numerical Solutions of 2-D Incompressible Driven Cavity Flow with Wavy Bottom Surface. American
Journal of Applied Mathematics. Special Issue: Fluid Flow and Heat Transfer Inside a Closed Domain. Vol. 3, No. 1-1, 2015, pp. 30-42.
doi: 10.11648/j.ajam.s.2015030101.14

Abstract: In the present numerical study is devoted to investigate the lid-driven cavity flow with wavy bottom surface. The
cavity upper wall is moving with a uniform velocity by unity and the other walls are no-slip. The physical problem is represented
mathematically by a set of governing equations and the developed mathematical model is solved by employing Galerkin
weighted residual method of finite element formulation. The wide ranges of governing parameters, i. e., the Reynolds number
(Re), and the number of undulations () on the flow structures are investigated in detail. The behavior of the force coefficient Cf
also has been examined. Streamline plots provide the details of fluid flow. The fluid contained inside a squared cavity is set into
motion by the top wall which is sliding at constant velocity from left to right and the undulation which was induced at the bottom
surface. It is found that these parameters have significant effect on the flow fields in the cavity. Furthermore, the trends of skin
friction for different values of the aforementioned parameters are presented in this investigation.

Keywords: Skin Friction; Lid Driven Cavity; Numerical Study, Wavy Surface

1. Introduction
Lid-driven flow in a two-dimensional square cavity is a
classical, benchmark problem, which involves complex flow
phenomena such as different types and scales of vortices,
bifurcation, transition and turbulence. These important flow
characteristics are closely relevant to a number of practical
applications [1, 2]. Driven cavity flow serve as a benchmark
problem for numerical methods in terms of accuracy,
numerical efficiency and etc. Many studies can be found on
the two-dimensional square cavity flow with its upper lid
sliding at a constant speed. In fact, the literature review
shows that two categories of studies were investigated [3].
The first category is concerned with a horizontal sliding lid,
which encompasses the top wall [48], bottom-sliding wall
[9], or an oscillating lid [1012]. The second category is
related with side driven differentially heated enclosures,
where one wall or both vertical walls move with a constant
speed [1315].
Due to the simplicity of the cavity geometry and regular
boundary conditions, applying a method on this flow
considerably easiness and simplification in the numerical
coding which allow to learn much about the fundamental
issues underlying the steady flows in a finite closed enclosure.
Despite its simple geometry, the driven cavity flow retains a
rich fluid flow physics manifested by multiple counter
rotating recirculating regions on the corners of the cavity
depending on the Reynolds number. Erturk et al [16] states
that a new quaternary vortex at the bottom left corner and a
new tertiary vortex at the top left corner of the cavity are
observed in the flow field as the Reynolds number increases.
Erturk and Gokcol [17] have investigated the 2-D steady
incompressible flow inside a triangular cavity. Their study
shows that the flow structures in a triangle cavity greatly
affected by the triangular geometry. Migeon et al. [18]
studied experimentally the effects of lid-driven cavity shape
on the flow establishment phase for square, rectangular and
semi-circular cavities.
Recently, Glowinski et al. [19] applied a finite element
method to the wall-driven flow in a semi-circular cavity and
revealed the vortex structure at high Reynolds numbers.
Mercan and Atalik [20] have studied the high-Reynolds
number lid-driven flow in arc-shape cavities with different
cross sections is considered up to Re = 8000. They have
 American Journal of Applied Mathematics 2015; 3(1-1): 30-42 31

discussed the effects of aspect ratio or arc angle, r, on the Mass conservation equation:
formation and growth of vertical structures, as well as on the
existence and development of periodic solutions. The works u v
+ =0 (1)
of Ostrach [21], Catton [22] and Yang [23] which described x y
the importance of the inclined cavity. Sheikholeslami et al.
[24] studied the problem of natural convection between a Momentum conservation equation:
circular enclosure and a sinusoidal cylinder. They concluded
that streamlines, isotherms, and the number, size and u u 1 p 2u 2 u
u +v = + ( 2 + 2 ) (2)
formation of the cells inside the enclosure strongly depend on x y x x y
the Rayleigh number, values of amplitude and the number of
undulations of the enclosure.
v v 1 p 2 v 2v
Most of the studies in the literature are concerned with the u +v = +( 2 + 2 ) (3)
square or rectangular cavity flows, although in applications, x y y x y
the cavities may be non-rectangular. There are few studies
dealing with flows in curved cavities driven by a moving lid. where is the density of fluid, u=(u(x,y),v(x,y)) the flow
A sinusoidal wall is chosen for the bottom surface with velocity vector, p(x,y) the pressure, x=(x,y) the Cartesian
different undulations. Therefore, in the present study, a coordinates and the kinematic viscosity of the fluid.
numerical simulation has been conducted to realize the The governing equations (1)-(3) are non-dimensionalized
influence of the wavy bottom surface on the flow structures using the following dimensional variables:
and skin friction in the square cavity.
x y u v p
X = , Y = ,U = ,V = ,P = (4)
2. Mathematical formulation L L U0 U0 U 0 2

A two dimensional lid driven cavity of height and length L Introducing the above non-dimensional scales into the
with wavy bottom surface is considered, as shown in Fig. 1. governing equations, we obtain the non-dimensional form of
The shape of the bottom wavy surface profile is assumed to the equations as follows:
mock following the pattern y = A(1 cos(2x)) , where A is
U V
the dimensionless amplitudeof the wavy surface and is the + =0 (5)
number of undulations. The upper wall of the cavity is X Y
allowed to move in its own plane at a constant speed unity
U U P 1 2U 2U
and the other walls have no-slip condition. In general, the U +V = + ( + ) (6)
cavity fluid is assumed to be Newtonian and incompressible, X Y X Re X 2 Y 2
unsteady and laminar flow. It is further assumed that viscous
dissipation is neglected in this study. With abovementioned V V P 1 2V 2V
U +V = + ( + ) (7)
assumptions, the governing equations for conservations of X Y Y Re X 2 Y 2
mass and momentum can be written as
U = 1,V = 0 for 0 X 1, Y = 1
U = V = 0 for 0 X 1, Y = A(1 cos(2X )) (8)
U = V = 0 for 0 Y 1, X = 0, X = 1

Where
The Re is the boundary
dimensionless non-dimensional Reynolds
conditions are: number given
where U 0 is the reference velocity, is the
U0L
as Re =

kinematic viscosity and L is the reference scale velocity.
The relation defines skin friction at the upper wall
u
y =1
y
C fu =
1
U 0 2
2
In non-dimensional form which becomes
2 U
C fu = Y =1
Re Y
Hence the total skin friction at the upper wall is
s
Fig. 1. Schematic diagram of the cavity and boundary conditions

D = C fu dX
0
32 K. M. Salah Uddin and Litan Kumar Saha: Numerical Solutions of 2-D Incompressible Driven Cavity Flow with
Wavy Bottom Surface

The relation defines skin friction at the wavy lower wall velocity profile for base grid and fine grid remain almost
u same. Therefore, we can consider base grid for computational
y = A(1 Cos(2x)) procedure.
y
C fl =
1
U 0 2
2
In non-dimensional form which becomes
2 U
Cf = Y = A(1 Cos ( 2 X ))
Re Y
Hence the total skin friction at the wavy lower wall
s
D = C f dX
0

3. Numerical Method
The governing equations along with the boundary
conditions are solved through the Galerkin weighted residual
finite element method. The formulation of this method and
computational procedure is well described by Taylor and
Hood [25] and Dechaumphai [26]. The computational
domain is discretized employing the uniform mapped mesh
grid system. Then the Galerkin weighted residual technique
is used to convert the nonlinear governing partial differential
equations into a system of integral equations that can be
solved numerically. The integration involved in each term of
these equations is performed by using Gausss quadrature
method, which leads to a set of non-linear algebraic
equations. These equations are then modified by imposing
boundary equations that is transferred into linear algebraic
equations by Newton-Raphson iteration. Finally, these linear
equations are solved by applying Triangular factorization
method.
4. Grid Refinement Study
A numerical method is said to be convergent if the solution
of the discretized equations tends to the exact solution as the
grid spacing tends to zero. Convergence of a non-linear
problem can be investigated by numerical tests, i.e. repeating
the calculation on a series of successively refined grids. If the
method is stable and all approximations used in the
discretization process are consistent, we will find that the
computation does converge to a grid-independent solution. A
grid that was 50% coarser and finer than the base case grid
was used to conduct the grid independency test. The
computations at base case conditions were repeated on this
refined grid and the solutions are compared. The numerical
results and its accuracy greatly influenced by the size of the
grid. In this study, we have adopted three types of grids,
namely coarse grid, base grid and fine grid to examine the
influence of grid resolution on the numerical results. The
coarse grid and fine grid are the 50% refinement of the base
case grid. The base case grid consists of 125125 grid points
where as the coarse grid consists of 7171 grid points and the
fine grid consists of 190190 grid points. From Fig. 2 we
observe that for coarse grid, velocity profile and skin friction
profiles are deformed from the base line curve but the
Fig. 2. (a) u-velocity profile along vertical line through the center of the
cavity (b) variation of skin friction at the bottom wall for various grid
point,=0, Re=103.
5. Results and Discussions
The characteristics of the flow fields in the lid-driven cavity
are examined by exploring the effects of the Reynolds number
Re, number of undulations, and amplitude of the wavy bottom
surface. Such field variables are examined by outlaying the
steady state version of the streamlines. The model validation is
an essential part of a numerical investigation. Hence, the
outcome of the present numerical code was benchmarked
against the numerical results of Poochinapan and Ching Mai
[28], which were reported for two-dimensional
incompressible flow in a lid-driven cavity. The comparison
was conducted while employing the following dimensionless
parameters: Re = 100, 5000. Excellent agreement was
achieved, as illustrated in Fig. 3(a) & 3(c), between our results
and the numerical results of Poochinapan and Ching Mai [28]]
for the streamlines inside the cavity. Further validation was
performed by comparing the streamlines contour in the
lid-driven square cavity between the present work and that of
Moshkin & Poochinapan [27] while employing Reynolds
number Re=1000 depicted in Fig. 3(b). Again, the comparison
strikes excellent agreement between both results as displayed
in Fig. 3(b). These validations boost the confidence in our
numerical code to carry on with the above stated objectives of
the current investigation.
 American Journal of Applied Mathematics 2015; 3(1-1): 30-42 33

Re = 1102 Re = 1103 Re = 5 103

(b) (c)
(a)

(d) (e) (f)

Fig. 3. Streamline contour with a comparison to Poochinapan and Mai[28]( (a)& (c)) Moshkin and Poochinapan [27](b), present result (d), (e) & (f).

(b)

(a)
34 K. M. Salah Uddin and Litan Kumar Saha: Numerical Solutions of 2-D Incompressible Driven Cavity Flow with
Wavy Bottom Surface

(d)
(c)

Fig. 4. Streamline contour at Re = 110 2 , A=0.05 for (a) = 0 (b) = 1 (c) = 2 (d) = 3

The effect of number of undulation on the streamline
contour is depicted in Fig. 4 at Re = 1 103 from which we
observed that undulation of the bottom surface affect the eddy
on the bottom right corner and bottom left corner. A small
eddy appears at right lower corner when undulation is not
employed at the bottom lower surface. However, employing
undulation at the bottom lower surface, a very small eddy
occurs on the left and right bottom corner. Also in Fig. 4(d),
we see that very small tertiary eddy occur in the left and right
bottom corner, which indicate that the increasing the number
of undulation affect the flow to the turbulent direction.
The streamlines contour are illustrated in Fig. 5 for various
undulations at Reynolds number, Re=1000. from which we
observed that for zero, first and second undulation large
recirculation in the clockwise direction occur in the bottom
right and bottom left corner also for second undulation tertiary
eddy occur in the bottom right corner. We see from Fig. 5(d)
for third undulation on the concave up region secondary eddy
formulate and a small tertiary eddy occur on the bottom right
and bottom left corner. The influence of number of
undulations on the streamline contour are demonstrated in
Fig.7 at Re=5000. It is observed from the figure that for first
undulation secondary and tertiary eddy occurs in the bottom
right and bottom left corner. From Fig. 6(c), we see that
secondary and tertiary eddy occur in the bottom left and
bottom right corner and a secondary eddy occur in the concave
up region. From Fig. 6(d) we see that due to the symmetric
property of flow in two dimensional recirculating flow in the
clock wise direction that occur in the concave up region
co-operate with the recirculating flow that occur in the bottom
right and bottom left corner.
 American Journal of Applied Mathematics 2015; 3(1-1): 30-42 35

(a) (b)

(c) (d)

Fig. 5. Streamline contour at Re = 1 103 , A=0.05 for (a) = 0 (b) = 1 (c) = 2 (d) = 3

The vortices contour for the amplitude of the wavy bottom
surface A=0.05 and various number of undulations =0, 1, 2, 3
is shown in Fig. 7. It was observed from Fig. 7 that vortices
clustered near the left and right corner of the upper wall also
some vortices occur in the upper half of the cavity. Also for
undulations, small vortices occur at the upper pick point of the
wave. The vorticity contour for the amplitude of the wavy
bottom surface A=0.05 and various number of undulations
number =0, 1,2,3 is shown in Fig. 8. From this figure, it was
observed that vortiicity decreased in the bulk of the cavity and
clustered near the boundary and the undulations of the bottom
surface affect the vortices.

(a) (b)
36 K. M. Salah Uddin and Litan Kumar Saha: Numerical Solutions of 2-D Incompressible Driven Cavity Flow with
Wavy Bottom Surface

(c) (d)

Fig. 6. Streamline contour at Re = 5 103 , A=0.05 for (a) = 0 (b) = 1 (c) = 2 (d) = 3

The vorticity contour for and for the amplitude of the wavy
bottom surface A=0.05 (a) =0 (b) =1 (c)=2 (d) =3 is
shown in Fig. 9. From which it was observed that the vortices
more clustered near the boundary and there is no vortices
occur in the bulk of the cavity. These figures show the vortices
that are formed in the flow field increases as the Reynolds
number increases. The v-component velocity profile for
various undulations along horizontal line through the center of
the cavity has been presented in Fig. 10. It is observed that for
low value of the Reynolds number, v-velocity along horizontal
line through the cavity slightly affect near the left and right
wall but for higher value of the Reynolds number effect of
undulations on the v-velocity profile decreases. The
u-component velocity profile for various undulations along
vertical line through the center of the cavity is depicted in Fig.
11. From which it is observed that for low value of the
Reynolds number there is not accountable effect on the
u-velocity but for higher value of the Reynolds number
u-velocity profile affected by undulations of the bottom
surface near the bottom surface and effect of undulation
decreases as the value of y increases.

(a) (b)
 American Journal of Applied Mathematics 2015; 3(1-1): 30-42 37

(c) (d)

Fig. 7. Vorticity contour for Re = 1102 , A=0.05 (a) = 0 (b) = 1 (c) = 2 (d) =3

(a) (b)

(c)
(d)

Fig. 8. Vorticity contour for Re = 1 103 , A=0.05 (a) = 0 (b) = 1 (c) = 2 (d) =3
38 K. M. Salah Uddin and Litan Kumar Saha: Numerical Solutions of 2-D Incompressible Driven Cavity Flow with
Wavy Bottom Surface

(a) (b)

(c) (d)

Fig. 9. Vorticity contour for Re = 5 103 , A=0.05 (a) = 0 (b) = 1 (c) = 2 (d) = 3

(a) (b)
 American Journal of Applied Mathematics 2015; 3(1-1): 30-42 39

(c)

Fig. 10. v-velocity profile for various undulations along horizontal line through the center of the cavity, A=0.05 (a) Re = 1 102 (b) Re = 1 103 (c) Re = 5 103 .

(b)
(a)

(c)

Fig. 11. u-velocity profile for various undulations along vertical line through the center of the cavity, A=0.05 (a) Re = 1102 ; (b) Re = 1103 ; (c)
Re = 5 103 .
40 K. M. Salah Uddin and Litan Kumar Saha: Numerical Solutions of 2-D Incompressible Driven Cavity Flow with
Wavy Bottom Surface

(a) (b)

(c) (d)

Fig. 12. Variation of skin friction along bottom surface (a) = 0 (b) = 1 (c) = 2 (d) = 3

(a) (b)
 American Journal of Applied Mathematics 2015; 3(1-1): 30-42
(c)
Fig. 13. Skin friction along wavy bottom surface for various undulation (a) Re = 1102 (b) Re = 1103 (c) Re = 5 103
Fig. 12 represents the variation of skin friction for Reynolds
number, Re= 1 10 , 1 10 , 5 10 . From Fig. 12(a) we
observed that negative skin friction for = 0 become
maximum near at the middle point of the cavity lower
boundary and it became minimum between the points x = 0.2
and x = 0.3. For = 1 negative skin friction profile look like
similar as for = 0 but the values of negative skin friction
grow for undulation at the pick point. From Fig. 12(c) and Fig.
12(d) we see that the negative skin friction rise depending on
the increasing of the curve and it became maximum at
maximum point but it decrease as the curve decrease become
minimum at local minimum point but for higher Re have some
complicated nature near the left and right boundary wall.
Fig. 13 represents the negative skin friction for various
values of Re. From Fig. 13(a) we observed that for skin
friction smoothly increases for =0 and it becomes maximum
the middle point of the lower surface and then decreases to
zero at the right boundary. But employing the number of
undulations at bottom surface the skin friction increases as the
surface increases. The highest skin friction is found at the
upper peak point. It decreases again when the curve goes
downward. Thereafter, the lowest skin friction is obtained at
the lower peak point. From Fig. 13(b) we observed that for flat
surface, the skin friction slightly decreases near the left
boundary and then increasing but the nature of the skin friction
remain same as for except the maximum value of skin friction
become lower. From Fig.13(c) we observed that it has the
same nature as for except the maximum and minimum value.
6. Conclusion
In the present work deals with the effects of undulated
bottom surface on a lid driven square cavity flow. In this paper
we conducted a simulation of incompressible viscous flow in a
lid-driven square cavity, described the cavity behavior of this
flow at the high Reynolds numbers, examined effects of Re,
number of undulations and the grid size on numerical
solutions of streamlines. The results obtained show that the
41
wall undulation affects the flow in the lid driven square cavity.
Employing the number of undulations at bottom surface, the
skin friction increases. The skin friction becomes highest at
the upper peak point and it becomes lowest at the lower peak
point.
References
[1] Shankar PN, Deshpande MD., Fluid mechanics in the driven
cavity, Ann Rev Fluid Mech 2000; 32:93136.
[2] Robert NM, Pavageau M, Rafailidis S, Schatzmann M. Study
of line source characteristics for 2-D physical modeling
ofpollutant dispersion in street canyons, Journal of Wind
Engineering and Industrial Aerodynamics, 1996; 62: 3756.
[3] R. Iwatsu, J.M. Hyun, Three-dimensional driven-cavity flows
with a vertical temperature gradient, Int. J. Heat Mass Transfer
38 (1995)33193328.
[4] A.A. Mohamad, R. Viskanta, Transient low Prandtl number
fluid convection in a lid-driven cavity, Num. Heat Transfer:
Part A 19(1991) 187205.
[5] A.K. Prasad, J.R. Koseff, Combined forced and natural
convection heat transfer in a deep lid-driven cavity flow, Int. J.
Heat Fluid Flow 17 (1996) 460467.
[6] C.F. Freitas, R.L. Street, Non-linear transient phenomena in a
complex recirculating flow: A numerical investigation, Int. J.
Num. Methods Fluids 8 (1988) 769802
[7] A.A. Mohamad, R. Viskanta, Flow and heat transfer in a
lid-driven cavity with stably stratified fluid, Appl. Math. Model
19 (1995) 465472.
[8] K.M. Khanafer, A.J. Chamkha, Mixed convection flow in a lid
driven enclosure with a fluid-saturated porous medium, Int. J.
Heat Mass Transfer 42 (1999) 24652481.
[9] C.-J. Chen, H. Nassari-Neshat, K.-S. Ho, Finite-analytical
numerical solution of heat transfer in two-dimensional cavity
flow, Num. Heat Transfer 4 (1981) 179197.
42 K. M. Salah Uddin and Litan Kumar Saha: Numerical Solutions of 2-D Incompressible Driven Cavity Flow with
Wavy Bottom Surface
[10] R. Iwatsu, J.M. Hyun, K. Kuwahara, Convection in a
differentially heated square cavity with a torsionally-oscillating
lid, Int. J. Heat Mass Transfer 35 (1992) 10691076.
[11] R. Iwatsu, J.M. Hyun, K. Kuwahara, Numerical simulation of
flows driven by a torsionally oscillating lid in a square cavity, J.
Fluids Eng. 114 (1992) 143149.
[12] Abdalla Al-Amiri, K. Khanafer, I. Pop, Numerical simulation
of unsteady mixed convection in a driven cavity using an
externally excited sliding lid, Eur. J. Mech. B/Fluids, in press
[13] V.S. Arpaci, P.S. Larsen, Convection Heat Transfer,
Prentice-Hall, 1984, p. 90.
[14] O. Aydn, Aiding and opposing mechanisms of mixed
convection in a shear-and buoyancy-driven cavity, Int. Comm.
Heat Mass Transfer 26 (1999) 10191028.
[15] A.J. Chamkha, Hydromagnetic combined convection flow in a
vertical lid-driven cavity with internal heat generation or
absorption, Num. Heat Transfer: Part A 41 (2002) 529546.
[16] E. Erturk, T. C. Corke2 and C. C. Gokcol, Numerical solutions
of 2-D steady incompressible driven cavity flow at high
Reynolds numbers, International Journal For Numerical
Methods In Fluids, 2005; 48:747774
[17] E. Erturk and O. Gokcol, Fine grid numerical solutions of
triangular cavity flow, The European Physical Journal Applied
Physics, 2007, 38, 97105
[18] C. Migeon, A. Texier, G. Pineau, Effects of lid driven cavity
shape on the flow establishment phase, J. Fluids Struct. 14
(2000) 469488.
[19] R. Glowinski, G. Guidoboni, T.W. Pan, Wall driven
incompressible viscous flow in a two-dimensional
semi-circular cavity, J. Comput. Phys. 216 (2006) 7691.
[20] H. Mercan, K. Atalk,Vortex formation in lid-driven arc-shape
cavity flows at high Reynolds numbers, European Journal of
Mechanics B/Fluids 28 (2009) 6171
[21] S. Ostrach, Natural convection in enclosures, in: J.P. Hartnett,
T.F. Irvine Jr. (Eds.), Advances in Heat Transfer, vol. 8,
Academic Press, New York, 1972, pp. 161227.
[22] I. Catton, Natural convection in enclosures, in: Proceedings of
the Sixth International Heat Transfer Conference, vol. 6, 1978,
pp. 1331.
[23] K.T. Yang, Transitions and bifurcations in laminar buoyant
flows in confined enclosures, J. Heat Transfer 110 (1988)
11911204.
[24] Sheikholeslami, M., Gorji-Bandpy, M., Pop, I. & Soleimani, S.
(2013). Numerical study of natural convection between a
circular enclosure and a sinusoidal cylinder using control
volume based finite element method. International Journal of
Thermal Sciences, Vol. 72, pp. 147-158.
[25] C. Taylor, P. Hood, A numerical solution of the NavierStokes
equations using finite element technique, Computers & Fluids
1 (1) (1973) 7389
[26] P. Dechaumphai, Finite Element Method in Engineering, 2nd
ed. Chulalongkorn University Press, Bangkok, 1999.
[27] N. P. Moshkin, K. Poochinapan, Novel finite difference scheme
for the numerical solution of two-dimensional incompressible
Navier-Stokes equations, International Journal Of Numerical
Analysis And Modeling, 7(2)(2010) 321329
[28] K. Poochinapan and Chiang Mai, Numerical Implementations
for 2-D Lid Driven Cavity Flow in Stream Function
Formulation, ISRN Applied Mathematics, 2012 (2012), pp. -
1-17.